Topic: rpietila Altcoin Observer - page 125. (Read 387493 times)

I haven't studied thermodynamics, but I really like to keep using the word 'entropy'.   

"I haven't studied the Boolberry PoW algorithm but I do invent Proof Of Troll"

Ahh, you tricked me into reading!

Sentences of the form:

"I haven't studied the Boolberry PoW algorithm but "

typically deserve to have the "but" and all words following it replaced by a period.

* I haven't studied medicine or electromagnetics, but I believe that cell phones cause cancer!

* I haven't studied mechanical engineering, but I believe my bridge design is sound!

Don't those just read a lot better as:  "I haven't studied mechanical engineering." ? :-)

can't you see this man is serious! stop wasting his goddam time and get bruce schneir on the case already

You apparently have no comprehension of what has been said.

"less predictable parts of blocks" makes no quantitative sense. That they were generated from prior PoW is not sufficient to quantify them as "less predictable". For one thing, they are already known by the time they are used. It is not even necessarily true that the PoW hashes couldn't contain planted patterns (remember any value less than the required difficulty is acceptable)!

The Cryptonote PoW algorithm runs the current value through a hash, and uses the output to as the index to lookup the next random memory location containing the next value.

If that hash is not uniformly distributed and or not perfectly random, then the memory locations visited may not comprise all locations in the scratchpad or can be gamed in other ways.

For example birthdays are uniformly distributed but if the sampling size is too small, then the test of duplicate birthdays probabilities are not uniform! So seemingly random and uniformly distributed data is not in another context. The size of the sample (the entropy) matters.

That is just one possible weakness. There may be others.

Using naked (unenveloped) AES rounds as a hash function can be incorrect. I cited a reference on that already.

I haven't studied the Boolberry PoW algorithm but I am aware it is using data from the block chain to modulate the choice of the next index in the scratchpad. The potential problem is that data may not have the degree of uniform distribution and randomness required. Your notion of "less predictable" is mathematical nonsense. It is the period of cyclic structure and extent of entropy that matter in Birthday attacks. The Boolberry PoW algorithm may be replacing the pseudorandom generator in Cryptonote entirely with data from the block chain. Since that data is known a priori, it might be possible to precompute certain lookup tables or other cryptanalysis strategies.

I believe it is possible to fix both if they have weaknesses (well at least Cryptonote but the Boolberry PoW might be doomed if I am correct that planted patterns can be put into block solution hashes). For example for Boolberry, he could probably add a hash computation every so many lookups in the scratchpad (within the loop inside the over PoW hash), to sufficiently randomize and disperse any accumulative effects from the block chain data. Ditto Cryptonote can probably replace the AES circuit with a known secure hash has every N lookups in the scratchpad. But I would prefer cryptanalysis to tell us with more certainty.

WHY ARE YOU WASTING MY TIME? Hire a cryptographer to do some study.

Gosh, ton of info, discussion here.

I looked into Boolberry and Monero.

They're both good(pushing away the gpu miner guy for bbr)

But really, Boolberries improvements on the coin, can be done on Monero or any other Cryptonote coin just as quickly(without full copy/paste)

I choose Monero over Boolberry (disregarding instamines or anything else, just focusing on first appearances), because Monero came out first, it has a big, juicy, sized developers team, and the name is just better.

Its like comparing "A Coin"  with "W Coin", "A Coin" came out first, and acquired a good size development team, its name is also better. It's development team has made more "underneath" changes to the coin than "W Coin" and its developers are focusing on addressing different issues correctly and properly, unlike "W Coin", which it's developers are hurrying to change this and that, to try and make it stand out from "A Coin"

Basically "A Coin"/Monero, on first view as I have said, seems like its development is more focused on fully understanding and addressing issues without hurrying through things and finding "quick ducttape solutions" that cause more problems down the road, and "W Coin"/Boolberry seems like its developers are more focused on adding on this and that without fully understanding the code and not taking their time.
That's my view.

As we can see with the trend in disregarding the fact that these things are supposed to be a currency, everyone wants to be the next Dogecoin.

No, we aren't concerned about miners. 90% of blocks are mined by public pools anyway, there's no real point. We're mostly concerned about the people using it for barter transactions.

Cute mascot, it is the next Doggiecoin.

+





?

Martin Armstrong talks about what is coming after Sept. 2015 from the 34 min point forward in the a July 14 interview, but the most interesting part for us might be from the 43 min point forward (especially 46:30) where he talks about alternative currencies during this coming period. You can quickly skip forward in the interview to the juicy part. His war and social unrest cycle will start firing zonkers starting this November.

Boolberry does sound funny and not at all like money. I wonder how they came up with it.

So mining isn't anonymous? 

The I2P network is designed for interaction of persons making transactions with main nodes without revealing your IP. It's not intended for general use (we will continue to use the normal P2P code for that).

With all due respect (and I have a few bbr in my pocket), this may be the biggest "glass house" statement I have ever read.

Monero is widely regarded a terrific name for a cryptocurrency.  Smells like money in many languages.  What does boolberry smell like?

Simply from marketing and brandability point of view (it's a vanilla scrypt clone), what do you think about iCoin? Mainstream potential?

I'm bummed its dev quit and took off months ago, so the coin is pretty much dead now. The remaining marketing and web design guys quit also soon after. They said they made sure they wouldn't get into trouble using the name iCoin, in case someone thinks Apple has trademarked i*.

The website seems to be down, but I found this short video displaying the logo design: http://www.youtube.com/watch?v=ygdXrS4XoWQ

Wouldn't mind if someone happened to design a perfect crypto coin and taking over iCoin brand and giving the current iCoin holders a few new coins in return.

Good. You've got a $6 million marketcap, surely you can rub a few nickels together to hire some cryptanalysis. You all said you inherited a crud PoW implementation, so it behoves you to not play el cheapo on such a fundamental aspect of a crypto-coin.

Heck I haven't been paid a penny, and the PoW I did has extensive cryptanalysis. You better be ready!


And you are respected for that.


But this childish melodrama is pathetic. And reflects badly on the coin you say you are trying to help.

I was never attacking the valuable work you did on cleaning up the crud in the implementation of the PoW. I merely pointed out that a non-uniform and or non-random distribution on the lookups in the scratchpad can enable reduction of scratchpad. Since the default Crytonote PoW algorithm is supposedly random memory latency bound, then smaller memory footprint might mean moving into L2 cache which has significantly lower (faster) latency.

Othe, according to things that smooth said about problems with nodes bandwidth overhead, this i2p implementation looks really strange, since it definitely make this also slower.

I have no doubts that you informed about all issues and would be happy to see a result when you finish that.

I mean we don't make any claims of it never having been broken or being perfectly secure. I'm not sure if you follow the Monero Missives at all, but when we released the whitepaper review we said the following:

+1 for deescalation by chuckle.